Negative velocity values during traverse

[szw21]

Can you get negative values while traversing a duct. Does this indicate localized air turbulence and therefore not a suitable location for a traverse. Or is that something that you may encounter in many if not most or all situations?

[heatingman]

It means there is turbulence/eddy currents

Believe it or not, Ive seen 6” supply take offs mounted in such a way that they actually flow in reverse due to the same phenomenon.


Sent from my iPhone using Tapatalk

[szw21]

Does that mean that the location I'm using is not suitable for a traverse. When everything is averaged out the final result looks close to what I'm expecting but I do see odd reading at some point during the process.

[WAYNE3298]

Your traverse is bogus. You need to find another spot.

[szw21]

Your traverse is bogus. You need to find another spot.

ok thanks. When you do a traverse do you keep an eye on each reading and if so what do you look for to tell you it might not be a good spot. Just for positive readings or something else. I believe ASHRAE 111 specifies something about the range of values but I cannot recall what it is offhand without looking at the standard offhand plus you won't be able to apply that rule until all the points have been collected.

[WAYNE3298]

For an acceptable traverse 75% of the readings have to be within 10% of the maximum reading. Negative readings are acceptable and are added to the total as zero but will bogus a traverse in a hurry.

The reason I told you your traverse was bogus with negative readings is that in the 25 years I did TAB I ran thousands of pitot traverses and never saw one that was good with negative readings because there were always too many of them. With a pitot I wanted an accuracy with a max error of 5% or less.

[szw21]

For an acceptable traverse 75% of the readings have to be within 10% of the maximum reading. Negative readings are acceptable and are added to the total as zero but will bogus a traverse in a hurry.

The reason I told you your traverse was bogus with negative readings is that in the 25 years I did TAB I ran thousands of pitot traverses and never saw one that was good with negative readings because there were always too many of them. With a pitot I wanted an accuracy with a max error of 5% or less.

OK thanks. Let me try again and see what I get.

[heatingman]

You need to measure in the straitest segment of duct possible.

Many times, especially on smaller systems or those in tight locations, there are no good locations, so you do the best you can.

Then confirm readings with another style of flow measurement.


Sent from my iPhone using Tapatalk

[Artrose]

Good on you for asking the question you asked. I'd say you are definitely in a bad traverse location. In medium pressure ductwork, it's not all that uncommon to find what you've found, especially in larger ductwork. Sometimes, depending on the flows, and the ductwork, you also might encounter very small Vp readings that are difficult to accurately record.

You are looking for an area in the duct that has laminar airflow, or as close to laminar flow as possible, and your equidistant reading points must follow the traverse standards.

There are a few rule of thumbs that go with acceptable traverse locations. One rule of thumb states that, if possible, your readings should be taken 3 to 4 duct diameters down stream from any source of disturbance. (take off, transition, etc.) Basically, take your readings in an area of the duct that is as free from as much disturbance as possible. That's where the art, science, and your experience, come into play.

I've watched dingleberries randomly poke a hole in a round duct, then find a spot with their pitot that suits them inside the duct, then guarantee the accuracy of their cfm reading. I've seen the same half-azz approach to rectangular ducts, and with square oval.......forget about it......., these jokers are totally clueless.

[WAYNE3298]

I had an engineer question a traverse because it showed the airflow on an AHU in a system he designed was low on airflow. I had a tough time finding a place for a good traverse and even drilled the duct near the fan. Almost all the readings were negative with my Shortridge so it was useless. The engineer used an analog manometer and traversed that spot and informed me at the meeting the AHU was not low on airflow. He insisted so all of us went to that spot to do a traverse. He lost that argument.
What I'm saying don't use an analog manometer.

[szw21]

I'm still in learning mode with the new Alnor EBT731 I purchased so I'm trying to make sense of my readings. I mostly do residential and it is not easy to find a section of duct which obeys the rules for a good traverse.

Playing around with my own home system to get a feel for it. I did a traverse out of the supply plenum and got one or two negative flows but when averaged the flow gives good correlation with the blower chart based on TESP and also by measuring the return flow using the hood. I'm using the log-Tchebycheff traverse pattern - the pitot tube that came with the system has little rings you can preset at the correct locations. I might make up an excel spreadsheet to generate the correct pattern or there is probably some online calculator for this.

[Artrose]

I'm still in learning mode with the new Alnor EBT731 I purchased so I'm trying to make sense of my readings. I mostly do residential and it is not easy to find a section of duct which obeys the rules for a good traverse.

Playing around with my own home system to get a feel for it. I did a traverse out of the supply plenum and got one or two negative flows but when averaged the flow gives good correlation with the blower chart based on TESP and also by measuring the return flow using the hood. I'm using the log-Tchebycheff traverse pattern - the pitot tube that came with the system has little rings you can preset at the correct locations. I might make up an excel spreadsheet to generate the correct pattern or there is probably some online calculator for this.

If you could find a friendly NEBB balancer, you might ask him for a set of the old paper round, and rectangular, duct traverse forms. The traverse instructions are written on the pages. All you need to perform an old fashioned traverse, are those forms, a marker, a tape measure, knife, a roll of black tape, 3/8 bit, drill, hole plugs, duct tape, ladder, pitot tube, proper size magnehelic, calculator, safety glasses, (maybe some cardboard), and practice. Oh, and a simple shortened formula........ SqRt of the Vp times 4005 times duct area in SqFt. And the knowledge to put it all together.......... i.e. Training. (start with the pencil and paper first, then work up to the electronic stuff) It'll make you smarter.

[RobertStephens]

The engineer asked us to calibrate the airflow stations used to bring ERV air to the RTUs. So we drilled holes upstream of the sensors, and we got numbers like this. We did our job, but I don't know if we can certify the numbers. The engineer hasn't said anything so far...

-248	255	535	1,372
-236	354	823	1,484
1,084	1,370	1,321	1,517
1,110	1,207	1,334	1,460

[hvacker]

When we think of air flows in a duct we think of a smooth column of air but that's not the case. Ducts have friction so air in contact with the duct near it's sides tends to tumble. The pitot readings too close to this turbulence will have unreliable numbers. That's where we space the test holes to give the best readings.
Negative readings or even outlets that draw in air can have different causes but the reason is not enough static pressure.
Slipping blower drives, plugged filters, anything causing less than design out put but sometimes the duct is too large for the blower to fully pressurize.
I've seen this on extended plenums where ducts aren't reduced after take offs allowing air to favor downstream outlets. Then anything that affects design output will affect take offs.
Ducts that are too large are often the victims of the 0.1 Ductalator.

Test locations can throw a you a loop also. Too close to the fan outlet, to a change in direction, a change in duct size, all that stuff.

[Artrose]

The engineer asked us to calibrate the airflow stations used to bring ERV air to the RTUs. So we drilled holes upstream of the sensors, and we got numbers like this. We did our job, but I don't know if we can certify the numbers. The engineer hasn't said anything so far...

-248	255	535	1,372
-236	354	823	1,484
1,084	1,370	1,321	1,517
1,110	1,207	1,334	1,460

If those are foot per minute readings? And you added and divided to get your mean average. And you knew your duct area? You might have something to start with there? But as presented, that bunch of numbers is worthless. You have nothing to certify. To calibrate an airflow sensor, depending on the type, and the control system, you normally will need to adjust a multiplier, or K factor, etc., in the control program. After calibration, the airflow station readout should match your traversed duct airflow total.

[szw21]

If you could find a friendly NEBB balancer, you might ask him for a set of the old paper round, and rectangular, duct traverse forms. The traverse instructions are written on the pages. All you need to perform an old fashioned traverse, are those forms, a marker, a tape measure, knife, a roll of black tape, 3/8 bit, drill, hole plugs, duct tape, ladder, pitot tube, proper size magnehelic, calculator, safety glasses, (maybe some cardboard), and practice. Oh, and a simple shortened formula........ SqRt of the Vp times 4005 times duct area in SqFt. And the knowledge to put it all together.......... i.e. Training. (start with the pencil and paper first, then work up to the electronic stuff) It'll make you smarter.

I took the NCI balancing class last year so I'm familiar with the principles. I have their class materials along with the procedures. I also have the stuff you have mentioned including sheet metal hole cutters, drill etc. And I'm also familiar with the formulas regarding static pressure, velocity pressure, total pressure etc. and how to go from that to the cfm. The theory and procedure are straightforward and not difficult to understand.

The main challenge is to interpret the results and understand if the results make sense or not. I guess if you do it enough times you get a feel for a good traverse vs one that is not.

The micromanometer from the Alnor flow hood does all the conversion from SP/TP to cfm, collects all the points and does the averaging etc.

[szw21]

The engineer asked us to calibrate the airflow stations used to bring ERV air to the RTUs. So we drilled holes upstream of the sensors, and we got numbers like this. We did our job, but I don't know if we can certify the numbers. The engineer hasn't said anything so far...

-248	255	535	1,372
-236	354	823	1,484
1,084	1,370	1,321	1,517
1,110	1,207	1,334	1,460

Do you think you have enough data points? According to my Alnor flow hood manual you should have at least 24 - 25 points - more for larger rectangular duct.

[szw21]

Interesting article in TAB journal on equal area vs log-Tchebycheff.

http://docshare02.docshare.tips/file.../100899450.pdf

This is a rather old edition so maybe there are newer articles and procedures on traverse patterns.

[WAYNE3298]

The equal area traverse is not as accurate as the log-Tchebycheff but gives almost as good results and is a lot faster. The number of readings varies with the duct size. If you run comparison traverses the results are normally very close. The numbers in post #13 are obviously from an equal area traverse and the traverse results were no good.

[Artrose]

The equal area traverse is not as accurate as the log-Tchebycheff but gives almost as good results and is a lot faster. The number of readings varies with the duct size. If you run comparison traverses the results are normally very close. The numbers in post #13 are obviously from an equal area traverse and the traverse results were no good.

This is why some form of mentorship, or apprenticeship, is so vitally important for this work. You can buy em books and send em to school, and they come back with lots of big words and and theory, but no practical experience.
JCI does this to their Technicians. They send them to school for a week, then expect them to be experts when they hit the jobsite the next week. If you don't already know the material before you attend the class, you're screwed. I've been there multiple times myself, and it can be frustrating as hell.

By the way........obviously? I'm still not 100% sure what those numbers were supposed to represent.